HAROLD SIMMONS BOOTHAND ALBERTA. SCHWARTZ
2662
1701. 68
boiling points is applicable to the ethyl and ipropyl trihalogenosilanes. TABLE I1 BOILINGPOIST DIFFEREXESOF ETHYLAND TRIHALOGEKO SILASES Compound
c.
B. p . ,
C2H5SiC13 C2HjSiCl2F C2H5SiCIE'; C2H5SiF3
97.9 6% 2 37.3 -4.2
i-C,H,SiCI, i-C3H7SiC12F i-C3H&3iCllj2 i-C3H,SiFz
119.4 84.0
48.8 15.7
[ C!OSTRIBUTIOS
Diff. 'I
r
-
t5a. I
35.0 31.0 35.4 35.2 33.1
FROM THE
Acknowledgments.-This investigatioii was carried out under the sponsorship of the Naval Research Laboratory. We are deeply indebted t o members of the staff of the Chemical Division &PROPYLfor valuable suggestions and encouragement during its progress. Av. Summary Methods are described for the preparation and purification of i-propyl trichloro-, dichlorofluoro-, 34.1 chlorodifluoro- and trifluorosilane. All of these compounds are new substances not previously mentioned in the literature. Certain of their physical constants and chemical properties are described. 34.5
CLEVELAND, OHIO
RECEIVED ACGUST ~~ 17, 194G
(15) Original manuscript received July 18, 1946.
MORLEYCHEMICAL LABORATORY, WESTERN RESERVEUNIVERSITY]
VI. Preparation and Fluorination of n-Butyl Trichlorosilgnel BY HAROLD SIMMONS BOOTHAND ALBERTA. SCHWARTZ The use of the alkyl chlorosilanes in the prepara- ratio of silicon tetrachloride to butylmagriesiutn bromide not kept well above about 1.5: 1. tion of silicones by hydrolysis and condensation was The final purification of the n-butyl trichlorosilane for has necessitated more adequate information on the purposes of analysis and determinatjion of physical contheir properties. Little has been known about stants was achieved in an electrically heated, liquid disthe alkyl halogeno silanes containing fluorine. tillation column with an inside diameter of 10 mm. packed for a length of 50 cm. with single turn glass helices and This article presents the results of continued equipped with a Whitmore-Luxlo total condensation, systematic s t ~ d ofy these ~ ~compounds ~ ~ ~ and ~ ~ variable ~ ~ take-off type head" containing a thermocouple their fluorination products. well. The middle fraction was redistilled and used for In this investigation n-butyl trichlorosilane (12- study. Fluorination of n-Butyl Trichlorosilane .--The apparatus C4H9SiC13) was prepared, and fluorinated by the and experimental technique and procedures used in the Swarts r e a c t i ~ n . ~The , ~ fluorination yielded n- fluorination were the same as that described by Booth and butyl dichlorofluorosilane (n-C4H9SiC12F),n-butyl Spessard.6 I n every experiment the quantity of antimony chlorodifluorosilane (n-C4H&ClF2), and n-butyl fluoride used was in excess of the amount needed to fluorinate completely all the n-butyl trichlorosilane present. trifluorosilane (n-CdHgSiFJ. The fluorination of n-kutyl trichlorosilane proceeds Experimental Preparation and Purification of %-Butyl Trichlorosilane.--The n-butyl trichlorosilane used in these studies was prepared by the Grignard synthesis as described by Bygdeng save that excess silicon tetrachloride ( 2 moles for each mole of butylniagnesium bromide) and rapid stirring were necessary in order to reduce to a minimum the yields of the di-, tri- and tetra-n-butylsilane derivatives, which tended to form in considerable amounts, particularly if the 11) From a thesis submitted b y Albert A. Schwartz t o t h e Gradua t e School of \Vestern Reserve liniversity, June, 18-15,in partial fulfilment of t h e requiremerits of t h e degree o f Doctor of Philosophy, and based upon work done in connection with a research project sponsored by t h e Naval Research Laboratory, Office of Research and Inventioiis, U. :i. S a v y Department. Publication delayed for security reasons. ( 2 ) (I) H. S . Dooth a n d Paul Carnell, THISJOURNAL,68, 2650 (1946). (3) (11) H. S . Booth a n d H. S. Halbedel, ibid., 68, 2652 (1946). (4) (111) H. S. B o o t h a n d W. F. Martin, ibid.,68,2655 (1946). (5) ( I V ) H. S. Booth a n d J. F. Suttle, ibid., 68, 2658 (1946). ( 6 ) (V) H. S. Booth a n d Dwight R. Spessard, ibid., 68, 2660 (1946)' (7) F. Swarts, Bull. i l c a d . R o y . Belg., 24, 309,474 (1892). ( 8 ) H. S. Booth and C. F. Swinehart, THISJ O U R N A L , 54, 4751 (1932); 67, I333 (1D3.7). (9) A. Bygdeu Bel., 44,204ti (1911).
sluggishly below about 70 . At any given temperature there appeared to be a minimum pressure a t which fluorination proceeded smoothly. This minimum pressure increased with increasing temperature. A study of the log of various preparations (Table I) indicates that perhaps the most significant factor in i n creasing the yields of the intermediate chlorofluorides is the use of the catalyst. When the catalyst was omitted, only traces of these materials were obtained; xhen i t was used, the yields were appreciably better. The formation of n-butyl dichlorofluorosilane appears to be favored by relatively low temperatures and pressures. For any given pressure, increasing the temperature reduced the yield of this substance. A t any givrn temperature an increase in pressure had the same effcct . Apparently the formation of n-butyl chlorofluorosilane is favored by somewhat higher temperatures and pressures than is the formation of the n-butyl dichlorofluorosilane. These results are contrary to what we might expect, but they niay be due to the effect of two factors which do not show up in the table, namely, rate of mixing of the reaction mixture and rate of addition of the antimony fluoride. In the last four experiments the stirring was much more rapid than in the first three. Furthermore, in the first three runs the antimony fluoride was fed into the
( I O ) I; C. \Vhitniore and A . R . L u x , T H I S J U I I K N A L , 54, 3 1 i X (1032). ( 1 1 ) A A. Murton, "Laboratory l'echnique i n Organic Chcmistry," hicGrdw-~IiIillCo., X e w Yurk, Pi. T., 1938,PI). 83-85.
Dec., 19-46
P R E P A R A T I O N AND
IJOG O F
Expt.
FLUORINATION OF Iz-B UTYL TR1CHLOKOSILhNh
TABLE I FLUORISATION OF n-BUTYL TRICHLOROSILANE UNDER
\VI. o f UuSiCl:,
W t . of SbFa,
!4.
g.
Catalyst (SbClr), ml.
Generator b a t h temp.,
c.
Generator pressure,
mm. Hg
DIFFEREXT
COXDITIONS
Diluent
(c-CaHaCI?), ml.
2663
7-Approx.
BuSiClzF
?' & yield of productsaBuSiClFz
BuSiFs
235 10 70-83 175-225 10 2 33 300 10 100-112 175-225 7 1 35 120 135 .. 118-125 175-225 Tracc Trace 33 I :38 I .5:1 10 113-117 :300-:3so 12.i 8 4 x .5 1 I92 196 20 108-112 375-425 165 6 3 35 ii 172 187 20 112-117 300-350 6 3 40 I 169 174 15 108-110 375-425 5 2 40 a The material unaccounted for by these figures consisted very largely of n-butyl trichlorosilane which distilled out of the reaction flask before it could be fluorinated. Small amounts of hydrogen chloride, silicon tetrafluoride, and other decomposition products were also obtained. I n each experiment there was a small loss of material which could not be accounted for. This was probably liquid which remained behind in the generator due to sorption by the antimony salts 1 2 :3 4
344 29 1
-
reaction mixture ,teadily and the fluorinatioii 15 a\ forced, whereas in the last four experiments the antimony fluoride was added very slowly and intermittently and no forcing of the reaction occurred. Purification of Fluorination Products.-After preliminary separation in the liquid type column described, the tri- and di-fluorides were redistilled several times in a gas type colurnn'2-'3 and the middle fractions used for further study. The monofluoride was similarly purified in the liquid type column. Analysis.-Chlorine was determined quantitatively, and the presence of carbon and silicon were verified qualitatively, as described by Booth and Suttle.6 Fluorine was determined by the method devised by Siegel.l* Determination of Physical Constants.-Molecular weights were determined by the Regnault method in the manner described by Germann.16 Freezing points were determined using the apparatus and techniques developed previously in this l a b o r a t 0 r y . ~ 6 . ~ ~ %-Butyltrifluorosilane was the only one of the four compounds studied which gave a definite reproducible freezing point, and even in this case great care had to be exercised due t o the tendency of liquid t o supercool before freezing occurred. This suDercooling was of the order of 15'. The other compounds studied (n-butyl trichlorosilane, n-butyl dichlorofluorosilane and n-butyl chlorodifluorosilane) do not appear t o possess definite freezing points. In each case as these substances were cooled, they became more and more viscous until finally a transparent glass was formed. If these glasses were cooled rapidly by liquid nitrogen and then kept a t that tcmperature for some time, a transition appeared to take place. The glasses would shatter with an audible report into a multitude of particles which appeared to possess no regular planes of cleavage. This transition apparently n-as accompanied by a change in volume as In one ease it shattered the freezing point tube and in another instance it cracked it. Attempti were made to obtain melting points for the glasses anti for the products which resulted from the transitions rlrscrihed above, but n o breaks in the temperature-time curvcs could he observed. Vapor pressures were measured by the static method as described by Booth and Halbedel,3 and Booth and Suttle.6 Curves graphed from the vapor pressure data1*are given in Fig. 1. Expressions of the usual form, log p( ,,,,,,.) = d / T B Tvere derived; thc value.; for A. and B are given in Table 11.
Physical and Chemical Properties.-%-Butyl trichlorosilane and its three fluorination products are all colorless, mobile liquids. The completely fluorinated compound fumes only slightly on exposure to the atmosphere, undergoing hydrolysis quite slowly. The trichlorosilane and the two chlorofluorosilanes, on the other hand, fume readily in moist air and hydrolyze much more rapidly; so rapidly, in fact, that although they have sweetish odors, these are difficult to detect as they are almost completely masked by the odor of the hydrogen chloride resulting from the hydrolysis. The n-butyl trifluorosilane possesses a sharp, suffocating odor which is distinctly different from those of the other three compounds. I t is interesting t o note that when the two chlorofluorosilanes are exposed to the atmosphere the only odor which can be detected is that of hydrogen chloride. This would seem to indicate that in these halosilanes the silicon-chlorine bond is more susceptible to hydrolysis than is the silicon-fluorine bond. Although all four compounds are hydrolyzed without ex treme difficulty, they are not soluble in water t o any appreciable extent. When they are added to water, two immiscible layers result with the hydrolysis occurring a t the interface. n-Butyl trichlorosilane and its fluorination products are soluble in petroleum ether, benzene, toluene and various
+
(121 5. F. Dulton. J . S U CChem. . I d . , 38,45T (1919). (131 H. S. Booth mil A . R. Bozarth, I d . Eng. Chem., 29, 470 (1937). (1.1) W. Siegel, Z . m g w . C h r m . , 42, 856 (1929). (15) A. F. 0 . Germann. J . Phys. Chem., 19, 437 (1915). (16) H S Booth a n d F C . Mericola. THIS J O U R N A L , 62, 640 (1940). (17) H. S. B o o t h a n d D. R. XIartin, ;bid., 6 4 , 2108 (1042). I 18) F < x
